Obstructive sleep apnea (OSA) consists of episodes of partial or complete closure of the upper airway that occur during sleep and lead to breathing cessation (defined as a period of apnea > 10 sec). Symptoms include restlessness, snoring, recurrent awakening, morning headache, and excessive daytime sleepiness. Diagnosis is based on sleep history and polysomnography. Treatment is with nasal continuous positive airway pressure, oral appliances, and, in refractory cases, surgery. Prognosis is good with treatment. Most cases remain undiagnosed and untreated and are often associated with hypertension, atrial fibrillation and other arrhythmias, heart failure, and injury or death due to motor vehicle crashes and other accidents resulting from hypersomnolence.
In at-risk patients, sleep destabilizes patency of the upper airway, leading to partial or complete obstruction of the nasopharynx, oropharynx, or both. When breathing is diminished but not absent, the condition is called obstructive sleep hypopnea.
The prevalence of OSA is 2 to 9% in adults; the condition is under-recognized and often undiagnosed even in symptomatic patients. OSA is up to 4 times more common among men and 7 times more common among people who are obese (ie, body mass index [BMI] > 30). Severe OSA (apnea-hypopnea index [AHI] > 30/h) increases the risk of death in middle-aged men.
Anatomic risk factors include obesity, an oropharynx “crowded” by a short or retracted mandible, a prominent tongue base or tonsils, a rounded head shape and a short neck, a neck circumference > 43 cm, thick lateral pharyngeal walls, or lateral parapharyngeal fat pads. Other identified risk factors include postmenopausal status, aging, and alcohol or sedative use. A family history of OSA is present in 25 to 40% of cases, perhaps reflective of heritable factors affecting ventilatory drive or craniofacial structure. The risk of OSA in a family member is proportional to the number of affected family members.
Acromegaly, hypothyroidism, and sometimes stroke can cause or contribute to OSA. Disorders that occur more commonly in patients with OSA include hypertension, stroke, diabetes, gastroesophageal reflux disease, nocturnal angina, heart failure, and atrial fibrillation or other arrhythmias.
Because obesity is a common risk factor for both OSA and obesity-hypoventilation syndrome (see Complications), the conditions frequently coexist.
Inspiratory efforts against a closed upper airway cause paroxysms of inspiration, reductions in gas exchange, disruption of normal sleep architecture, and partial or complete arousals from sleep. These factors may interact to cause the characteristic symptoms and signs, including hypoxia, hypercapnia, and sleep fragmentation.
OSA is an extreme form of sleep-related upper airway resistance. Less severe forms that do not cause O2 desaturation include
People with upper airway resistance syndrome are typically younger and less obese than those with OSA, and they complain of daytime sleepiness more than do those with primary snoring. The symptoms, diagnostic evaluation, and treatment of snoring and upper airway resistance syndrome are otherwise the same as for OSA.
Symptoms and Signs
Although loud disruptive snoring is reported by 85% of OSA patients, most people who snore do not have OSA (see see Snoring). Other symptoms of OSA may include choking, gasping, or snorting during sleep, restless and unrefreshing sleep, and difficulty staying asleep. Most patients are unaware of these symptoms (because they occur during sleep) but are informed of them by bed partners, roommates, or housemates.
When awake, patients may experience hypersomnolence, fatigue, and impaired concentration. The frequency of sleep complaints and the degree of daytime sleepiness do not correlate well with number of nocturnal arousals.
The diagnosis is suspected in patients with identifiable risk factors, symptoms, or both. Criteria for diagnosis consist of daytime symptoms, nighttime symptoms, and sleep monitoring that documents > 5 episodes of hypopnea and/or apnea per hour. Specifically, in regard to symptoms, there should be ≥ 1 of the following:
The patient and any bed partners, roommates, or housemates should be interviewed. The differential diagnosis of excessive daytime sleepiness is broad (see Insomnia and Excessive Daytime Sleepiness (EDS)) and includes
An extended sleep history should be taken in all patients who
Most patients who report only snoring, without other symptoms or cardiovascular risks, do not need an extensive evaluation for OSA.
The physical examination should include evaluation for nasal obstruction, tonsillar hypertrophy, and pharyngeal structure and identification of clinical features of hypothyroidism and acromegaly.
The diagnosis is confirmed with polysomnography (see Testing), which includes continuous measurement of breathing effort by plethysmography, airflow at the nose and mouth using flow sensors, O2 saturation by oximetry, sleep architecture by EEG, chin electromyography (looking for hypotonia), and electro-oculography to assess the occurrence of rapid eye movements. Polysomnography records and helps classify stages of sleep and the occurrence and duration of apneic and hypopneic periods. The patient is also observed by video, and ECG monitoring is used to determine whether arrhythmias occur in conjunction with the apneic episodes. Other variables evaluated include limb muscle activity (to assess nonrespiratory causes of sleep arousal, such as restless legs syndrome and periodic limb movements disorder) and body position (apnea may occur only in the supine position).
The common summary measure used to describe respiratory disturbances during sleep is the apnea-hypopnea index (AHI), which is the total number of episodes of apnea and hypopnea occurring during sleep divided by the hours of sleep time. AHI values can be computed for different sleep stages. The respiratory disturbance index (RDI) is a similar measure, describing the number of times per hour that blood O2 saturation falls > 3% but also includes RERAs. If EEG monitoring is used, an arousal index (AI) can be computed, which is the number of arousals per hour of sleep. The AI may be correlated with AHI or RDI, but about 20% of apneas and desaturation episodes are not accompanied by arousals, or other causes of arousals are present. An AHI > 5 is required for the diagnosis of OSA; a value > 15 indicates a moderate level of sleep apnea, and a value > 30 indicates a severe level of sleep apnea. Snoring loudly enough to be heard in the next room confers a 10-fold increase in the likelihood of having AHI > 5. The AI and RDI correlate only moderately with a patient's symptoms.
Portable diagnostic tools are being used more often to diagnose OSA. Portable monitors can measure heart rate, pulse oximetry, effort, position, and nasal airflow to provide estimates of respiratory disturbances during self-reported sleep, thereby providing a value for AHI/RDI. Portable diagnostic tools are often used in combination with tools that calculate patients' risk (the sensitivity and specificity of the test depend on pretest probability). When portable tools are used, coexisting sleep disorders (eg, restless legs syndrome) are not excluded. Follow-up polysomnography may still be needed.
Measurement of thyroid-stimulating hormone can be done based on clinical suspicion. No other adjunctive testing (eg, upper airway imaging) has sufficient diagnostic accuracy to be recommended routinely.
Prognosis is excellent if effective treatment is instituted.
Untreated or unrecognized OSA can lead to cognitive impairment as a result of sleeplessness, which, in turn, can lead to serious injury or death caused by accidents, especially motor vehicle crashes. Sleepy patients should be warned of the risks of driving, operating heavy machinery, or engaging in other activities during which sleep attacks would be hazardous.
Adverse effects of hypersomnolence, such as loss of employment and sexual dysfunction, can affect families considerably.
Long-term cardiovascular sequelae of untreated OSA include poorly controlled hypertension, heart failure, and atrial fibrillation (even after catheter ablation) and other arrhythmias.
In addition, perioperative complications, including cardiac arrest, have been attributed to OSA, probably because anesthesia can cause airway obstruction after a mechanical airway is removed. Patients should therefore inform their anesthesiologist of the diagnosis before undergoing surgery and should expect to receive continuous positive airway pressure (CPAP) when they receive preoperative drugs and during recovery.
The aim of treatment is to reduce episodes of hypoxia and sleep fragmentation; treatment is tailored to the patient and to the degree of impairment. Cure is defined as a resolution of symptoms with AHI reduction below a threshold, usually 10/h.
Treatment is directed first at risk factors and then at OSA itself. Specific treatments for OSA include CPAP, oral appliances, and airway surgery.
Control of risk factors:
Initial treatment aims at optimal control of modifiable risk factors, including obesity, alcohol and sedative use, hypothyroidism, acromegaly, and other chronic disorders. Although modest weight loss (15%) may result in clinically meaningful improvement, weight loss is extremely difficult for most people, especially those who are fatigued or sleepy. Bariatric surgery frequently reverses symptoms and improves AHI in morbidly obese (BMI > 40) patients; however, the degree of these improvements may not be as great as the degree of weight loss. Weight loss, with or without bariatric surgery, should not be considered a cure for OSA.
Nasal CPAP is the treatment of choice for most patients with OSA and subjective daytime sleepiness; adherence is lower in patients who do not experience sleepiness. CPAP improves upper airway patency by applying positive pressure to the collapsible upper airway segment. Effective pressures typically range from 3 to 15 cm H2O. Disease severity does not correlate with pressure requirements. Many CPAP devices monitor CPAP efficacy and titrate pressures automatically, according to internal algorithms. If clinical improvement is not apparent, CPAP efficacy should be reviewed and patients should be reassessed for a second sleep disorder (eg, upper airway obstruction) or a comorbid disorder. If necessary, pressure can be titrated manually during monitoring with repeat polysomnography. Regardless of improvement in the AHI, CPAP will reduce cognitive impairment and BP. If CPAP is withdrawn, symptoms recur over several days, though short interruptions of therapy for acute medical conditions are usually well tolerated. Duration of therapy is indefinite.
Failures of nasal CPAP are common because of limited patient adherence. Adverse effects include dryness and nasal irritation, which can be alleviated in some cases with the use of warm humidified air, and discomfort resulting from a poorly fitting mask.
CPAP can be augmented with inspiratory assistance (bilevel positive airway pressure) for patients with obesity-hypoventilation syndrome (see Complications) to increase their tidal volumes.
Oral appliances are designed to advance the mandible or, at the very least, prevent retrusion with sleep. Some are also designed to pull the tongue forward. Use of these appliances to treat both snoring and OSA is gaining acceptance. Comparisons of appliances to CPAP show equivalence in mild to moderate OSA, but results of cost-effectiveness studies are not available.
Surgical correction of upper airway obstruction caused by enlarged tonsils and nasal polyps should be considered. Surgery for macroglossia or micrognathia is also an option. Surgery is a first-line treatment if anatomic encroachment is identified. However, in the absence of encroachment, evidence to support surgery as a first-line treatment is lacking.
Uvulopalatopharyngoplasty (UPPP) is the most commonly used procedure. It involves resection of submucosal tissue from the tonsillar pillars to the arytenoepiglottic folds, including resection of the adenoids, to enlarge the upper airway. Equivalence with CPAP was shown in one study using CPAP as a bridge to surgery, but the interventions have not been directly compared. UPPP may not be successful in patients who are morbidly obese or who have anatomic narrowing of the airway. Moreover, after UPPP, recognition of sleep apnea is more difficult because of a lack of snoring. Such silent obstructions may cause apneic episodes as severe as those occurring before surgical intervention.
Adjunctive surgical procedures include midline glossectomy, hyoid advancement, and mandibulomaxillary advancement. Mandibulomaxillary advancement is sometimes offered as a 2nd-stage procedure if UPPP is not curative. The optimal multistage approach is not known.
Tracheostomy is the most effective therapeutic maneuver for OSA but is done as a last resort. It bypasses the site of obstruction and is indicated for patients most severely affected (eg, those with cor pulmonale).
Laser-assisted uvuloplasty, uvular splints, and radiofrequency tissue ablation have been promoted as treatments for loud snoring in patients without OSA. Although they may transiently decrease snoring loudness, efficacy declines over months to years.
Adjunctive treatments are commonly used but have no proven role as first-line treatment.
Modafinil can be used for residual sleepiness in OSA in patients who are effectively using CPAP.
Supplemental O2 improves blood oxygenation, but a beneficial clinical effect cannot be predicted. Also, O2 may provoke respiratory acidosis and morning headache in some patients.
A number of drugs have been used to stimulate ventilatory drive (eg, tricyclic antidepressants, theophylline) but cannot be routinely advocated because of limited efficacy, a low therapeutic index, or both.
Nasal dilatory devices and throat sprays sold OTC for snoring have not been studied sufficiently to prove benefits for OSA.
Patient education and support:
An informed patient and family are better able to cope with a treatment strategy, including tracheostomy. Patient support groups provide helpful information and effectively support timely treatment and follow-up.
Last full review/revision March 2013 by Kingman P. Strohl, MD
Content last modified November 2013